BI Basics – Components of a Business Intelligence Process
Components of a BI Process
There are few phases or components which are common in any typical BI Process. Following are the three major components of any typical BI process:
- Design and Data Modeling
- Extraction, Transformation, and Loading
- Analytics and Reporting
In the above list, I have omitted few obvious phases/components like Requirements Gathering, Analysis, and Deployment etc. as they are common to any kind of project and are not specific to a BI Project. We will use the terms BI Process and BI Project interchangeably for the rest of this blog post.
Design and Data Modeling
This is the crucial phase of a BI Process.as this phase forms the base of the BI System’s Architecture and is the deciding factor for the outcome of the BI Project.
This phase requires a strong knowledge and expertise on Architecture and Design of Data Warehousing and Business Intelligence Solutions, Data/Dimensional Modeling, and the ability to look at a bigger and holistic picture while designing the solution.
Apart from these skills, one should also have complete understanding of the needs of the Business Users, Pain Points, and also expertise with tools like ERwin Data Modeler etc.
Extraction, Transformation, and Loading
This is the next phase of the BI Projects. Extraction, Transformation, and Loading is known as “ETL” in short. This phase involves consolidation of data from varied source systems, bringing them under one single set of standards/conventions, transforming the data by applying necessary business rules, and loading it into appropriate layers of a BI System (Like Staging, Warehouse, Data Mart etc.).
This phase requires a good understanding of different sources involved, BI system/Warehouse schema, and expertise on ETL tools.
Analytics and Reporting
This is the last phase of a BI project. Depending upon the business needs, this phase can be sometimes split into two phases as “Analytics” and “Reporting”.
In Analytics usually multi-dimensional structures, usually known as “Cubes”, are built if you are working in the Microsoft BI (SQL Server BI) space. And in reporting, various types of reports are created using one or more tools like SQL Server Reporting Services (SSRS), Power View, and Performance Point Services etc. or any other Analytics and Reporting tools available in the market.
Basically this entire phase deals with delivering the data to the end-user in the format which they like, understand, and on which they can act. This phase requires a strong knowledge of the end users habits of how they use the data, how they interpret the data, what kind of formats they use, and what kind of tools they are comfortable using etc. so that the data/information can be delivered in desired format with the help of desired tools at the right time to the decision makers.
There are other aspects like training, which is very critical to ensure that the system is adopted by larger audience and decision makers can use the solution to make effective decisions. However, this aspect is common to any other type of project.
There is another concept called “Self-Service BI” which is totally different and the above components are not really part of this concept. “Self-Service BI” deals with doing BI on our own, though there are many instances where some of the above processes are applicable like building a Warehouse, Consolidation of data etc. However, in many cases, end users do the reporting on their own.
Hope this gives a brief idea of what a typical BI process/project looks like.
Until next time, happy learning!
Posted on June 18, 2013, in Basic Concepts, SQL Server and tagged Analytics, BI Project, Business Intelligence, Data Mart, Data Warehousing, Databases, ETL, Extract transform load, Microsoft SQL Server, MSBI, Reporting Services, SQL Server, SSAS, SSIS, SSRS. Bookmark the permalink. 4 Comments.